C-peptide has been reported to prevent ocular vascular dysfunction in rats with diabetes of 5 weeks duration. This study was undertaken to compare effects of C-peptide alone, low dose insulin alone, and the combined treatments in reversing vascular dysfunction after 2 months of untreated diabetes. Male Sprague Dawley rats 270-300 g bw were made diabetic by i.v. injection of 50 mg/kg streptozotocin. Non-diabetic control received only buffer. After 2 months of diabetes, rats were randomly assigned to 4 groups (n=10 each) untreated diabetes (D), synthetic C-peptide (DC), insulin (DI) and C-peptide+insulin treatment (DCI). 400 ug/kg bw of human C-peptide was injected s.c. twice a day, and 10 U/kg bw of NPH insulin was injected once a day. One month later albumin permeation was assessed by using 125I- and 131I-albumin and blood flow was assessed by using 46Sc-microspheres. Body weight was increased by insulin, but not by C-peptide alone. Neither C-peptide nor insulin improved plasma glucose or glycated Hb. Diabetes increased albumin permeation in retina, anterior uvea and posterior uvea. Insulin or C-peptide alone did not affect albumin permeation. C-peptide+insulin decreased albumin permeation in retina, anterior uvea, and posterior uvea. Blood flows in D did not differ from controls, and were unaffected by C-peptide or insulin. Neither low dose insulin nor C-peptide alone were as effective as the combination in reversing diabetes-induced vascular dysfunction. These observations suggest a new use for C-peptide in the treatment of diabetic retinopathy.